A. Field of Invention
The present invention relates to an improved process for producing phenyl- or substitutedphenylalkanolamine compounds having pharmacologic activity and to novel intermediates useful in the process, and more particularly relates to a process for directly obtaining the desired (+)- or (-) enantiomer in essentially 100% ee (enantiomeric excess) without the need for tedious resolutions.
Many biologically active compounds and medicinals are synthesized as racemic mixtures. However, commonly, only one of the optical isomers has the desired properties, while the other may possess only very weak or a different, undesired pharmacological activity, or at worst, is toxic. This problem is best illustrated by the problems associated with Thalidomide which was unfortunately marketed as a racemic mixture of the toxic isomer, responsible for the well-publicized birth defects, as well as the active optical isomer which was free from the teratological side effects. The Thalidomide tragedy could have been avoided had there been a simple, economical process available for separating the isomers. Since then, the pharmaceutical industry has employed tedious and expensive resolution processes to insure that only the desired optical isomer is present in the finished formulation. It is therefore highly desirable for the pharmceutical industry to be able to obtain one enantiomer in a simple, direct, less costly process.
Many pharmaceutically active compounds are phenyl- or substituted-phenylalkanolamines having the basic structure ##STR1## wherein n is 1 or 2, R.sup.1 is hydrogen, acetoxy, phenyl or substituted phenyl and R.sup.2 is lower alkyl, phenyl or substituted phenyl.
Typical drugs of this kind include Isoproternol, Colterol, phenylephrine, Bitolerol, Dipeverfrin, and the major new anti-depressent agents, Tomoxetine, Fluoxetine and Nisoxetine.
Tomoxetine, [[R]-(-)-N=methyl-2-(2-methylphenoxy)-benzenepropamine hydrochloride, Eli Lilly and Company, LY 139603] is a new drug currently undergoing investigation as an antidepressant (R. L. Zerbe et al., J. Pharmacol. Exp. Ther. 1985, 232, 139). The (-)- optical isomer has been shown to be nine times more potent than the (+)- isomer. Unlike chemical tricyclic antidepressants such as imipramine, (-)-Tomoxetine has been shown to inhibit specifically norepinephrine uptake in humans at dosages which are clinically well tolerated and to be a relatively weak ligand for .sup..alpha. -1, .sup..alpha. -2 and .beta.-andrenergic receptors. The latter receptors are generally regarded as responsible for undersirable side-effects associated with antidepressants.
The patent literature preparation of (-)-Tomoxetine involves a long and tedious procedure culminating in a highly inefficient resolution (20%) of the racemic mixture. See Molly et al. U.S. Pat. No. 4,018,895 and Foster et al. Eur. Patent No. 0052492. Clearly, an enantiomeric preparation of (-)-Tomoxetine is needed. The best procedure to date provides an overall yield of the (-)- isomer of 14% and an optical purity of only 88% ee. The present invention provides a simple synthesis of both (-)-Tomoxetine, (+)-Tomoxetine is essentially 100% ee, as well as both optically pure enantiomers of the cognant compounds, Fluoxetine and Nisoxetine.
In general, many of these valuable drugs in addition to the Lilly antidepressants discussed above are synthesized via the Mannich reaction to produce the amino substituted arylalkyl ketone. Reduction of the ketone gives the alcohol. The mixture of enantiomeric arylalkanolamines are then resolved in a generally tedious, costly inefficient process.
Thus, there exists a need for a more efficient, cost effective, simple process for directly obtaining the desired isomer of Tomoxetine and other arylalkanolamine pharmaceutical agents in essentially 100% enantiomeric excess. The present invention provides such a process as well as valuable intermediates useful in such process.
B. Prior Art
Recently, it was discovered that diisopinocampheylchloroborane, hereafter Ipc.sub.2 BCl, derived from either (+)- or (-)-alpha-pinene is capable of reducing arylalkyl ketones to the corresponding alcohols in a highly enantioselective fashion. (Brown et al., J. Org. Chem. 1985, 50, 5446. See also Herbert C. Brown copending U.S. patent application Ser. No. 902,175 filed Aug. 29, 1986). The alcohols so obtained were simple compounds, containing no other functionality, and as such were end-products in themselves. However, for the construction of more complex molecules, often required of biologically active compounds, additional functionalities that can be further transformed are desirable. The halides are a particularly appealing functional group.
Chiral haloalcohols have been prepared in variable enantiomeric excess with a reagent developed by Itsuno et al. (J. Chem. Soc. Perkin Trans. I. 1985, 2615). In addition to inconsistent optical yields, the nature of the reagent remains in doubt. Soai et al. (J. Chem. Soc. Chem. Comm. 1986, 1018) have demonstrated that a chirally modified lithium borohydride can reduce beta-halogenoketones in 81-87% ee. More consistent results have been obtained for asymmetric reduction of 2-haloacetophenones with neat Alpine-Borane (Brown et al., J. Org. Chem. 1985, 50, 1384).
In my copending application Ser. No. 902,175, filed Aug. 29, 1986, I disclosed that diisopinocampheylhaloboranes (Ipc.sub.2 BX) are exceptionally effective asymmetric reducing agents for simple phenylketones to the corresponding alcohol. The reagents Ipc.sub.2 BX appear to be the most effective currently available for such asymmetric reductions, and one would expect that reduction of the arylalylketoamines should provide a direct route to the desired optically pure arylalkylaminoalcohols. Unfortunately, the presence of the amino groups prevents the desired reaction. Tertiary amines coordinate with Ipc.sub.2 BX to prevent its use for reduction. Primary and secondary amines react to give the amino-substituted boron compounds RNHBIpc.sub.2 +HX.
The present invention provides a solution to that problem by providing a method in which the Ipc.sub.2 BX reagents are used to reduce a halo-substituted arylalkylketone. Treatment of the haloaralkylalcohol with the appropriate amine gives the desired optically pure arylalkylaminoalcohol.
The present invention also fulfulls the need for a simple, reliable method of preparing each enantiomer of structurally diverse, optically pure haloalcohols at will, which are intermediates in the preparation of optically pure phenoxyhalides which are useful as intermediates in the synthesis of the optically pure, desired enantiomer of a pharmaceutically active phenyl- or substituted-phenylalkanolamine.